172 CARNEGIE INSTITUTION OF WASHINGTON. 



inferred, and with it the relative expansion coefficients of the two pieces of 

 glass, or of a piece of glass and an included fragment of glass, of crystallized 

 material, such as a stone, or of a metal wire. The disturbing effects of irregu- 

 lar boundary surfaces can be largely eliminated by immersion of the glass, 

 samples in a liquid of the same refractivity. 



(419) On tracing rays of light through a reflecting prism with the aid of a meridian projec- 



tion plot. F. E. Wright. J. Opt. Soc. Amer., 5, 410-419 (1921). (Papers 

 on Optical Glass, No. 37.) 



Prisms are used in optical instruments to change the courses of the paths of 

 transmitted rays of light ; prisms are of different shapes and sizes and are used 

 either singly or in combination. In many prisms it is difficult to visualize 

 the paths followed by the rays on transmission without graphical aid of some 

 kind. In the present paper the angle meridian projection plot is suggested aa 

 suitable for the purpose. Its apphcation to the study of prisms is illustrated 

 by a series of examples from the more common types of reflecting prisms. 



(420) The latent heats of fusion of nickel and monel metal. Walter P. White. Chem. and 



Met. Eng., 25, 17-21 (1921). 



By the common method of dropping from a furnace into a calorimeter, the 

 latent heats of nickel and of monel metal have been determined and their 

 specific heats for the intervals to 1360° and 1260°, respectively. The metals 

 were protected against appreciable oxidation by sealing into siUca-glass 

 containers. Some improvements were made in the technic of dropping the 

 materials from the furnace into the calorimeter. The precision of the tem- 

 perature measurement was studied and the results indicate a final accuracy of 

 better than 1 per cent. This indication tends to be confirmed by the agree- 

 ment of all the determinations but one, which was 1.3 per cent divergent. 

 Several determinations were made by pouring directly into the calorimeter 

 nickel just ready to sohdify, giving agreement to about 2 per cent with the 

 furnace results. The latent heat of nickel, 73 calories per gram, is 17 calories 

 greater than the recent determination of Wiist. The latent heat of monel, 

 68 calories, is in excellent agreement with that of nickel. 



(421) Aphthitalite from Kilauea. H. S. Washington and H. E. Merwin. Am. Mineral- 



ogist, 6, 121-125 (1921). 



This paper describes aphthitalite collected in September 1920 from a hot 

 crack in a recent lava flow of the eruption of 1919-20 of Kilauea. The mineral 

 is uniaxial and rhombohedral, and is compared with occurrences at Etna, 

 Searle's Lake, and elsewhere. Chemical analysis shows that the ratio K2SO4: 

 Na2S04 is about 1 : 1.5, like that at Etna, while the ratio at Vesuvius and 

 Searle's Lake is about 3:1. It is shown that there is an isodimorphous series 

 from arcanite (K2SO4) to thenardite (Na2S04), both crystallographically and 

 optically. The presence of CUSO4 in the mineral is discussed, and the iso- 

 morphism of this is indicated. The mode of origin of the mineral is suggested 

 as the oxidation of vaporized sulphides of the alkalies and copper. 



(422) Dispersion in optical glasses: III. F. E. Wright. J. Opt. Soc. Amer., 5, 389-397 



(1921). (Papers on Optical Glass, No. 38.) 



In this paper a reference table of dispersions in optical glasses is given; also 

 a table of differences between the values in the reference table and the corre- 

 sponding values of the glasses listed by Parra Mantois. The data prove that 

 the chief optical differences in glasses are not differences in the character of 

 the dispersion, but lie in the fact that two glasses may have the same disper- 

 sion relations and yet have appreciably different refractive indices. In the 

 dispersion formula both refringence and dispersion should be specifically 



